1. Have linear scoring methods :
This article might contradict my argument but adopting linear scoring methods while awarding marks for academic performance would be better in the context that a person with 81% would not have any significant advantage than a person with 79%. This would apply to b-schools like IIM-Ahmedabad. It is agreed that the IIMs need students who have been performing well throughout in their academics. But, there are two factors that they discount.
- A student with 79% is just 2% underperforming than a student with 81% and he needs to get 2% lesser opportunity in making it to that IIM.
- A student from commerce background might find it increasingly difficult to get an 80+ in his graduation because the total number of people who get that score would be far lesser than that of other streams. This could be applied for X and XII boards too where a student from AP board would have a definite advantage from a student from a Bihar board because the average percentages in AP is far higher than that in Bihar.
For example, the cutoff for General category student with Zero Work ex and a percentage of 80.1 is around 99.3 where as for the same student with a percentage of 79.9 is 99.98. And we had live cases discussed in one of the forums.
So, having linear scoring methods would make it fair for the border cases. This is because, a 2/3 or a 3/3 would not have a significant impact but when multiplied with a number like, it would make a huge difference. Click here to know the criteria adopted by various IIMs.
According to this method, the following would be the scoring pattern:
| Score in X/XII/graduation | Marks awarded (out of 3) |
| 80+ | 3 |
| 79-79.9 | 2.9 |
| 78-78.9 | 2.8 |
| 77-77.9 | 2.7 |
And so on.
With the new scoring method, the following would be the change in the scores that a student gets out for various cases:
| S. no | Score of a person in 10th class (A) | Score of a person in 12th class (B) | Score of a person in graduation (C) | Total score out of 27 according to earlier method (A*B*C) | Total score according to the new scoring method |
| 1 | 79 | 79 | 79 | 8 | 24.389 |
| 2 | 81 | 79 | 79 | 12 | 25.23 |
| 3 | 79 | 81 | 79 | 12 | 25.23 |
| 4 | 79 | 79 | 81 | 12 | 25.23 |
| 5 | 81 | 81 | 79 | 18 | 26.1 |
| 6 | 81 | 79 | 81 | 18 | 26.1 |
| 7 | 81 | 81 | 79 | 18 | 26.1 |
| 8 | 81 | 81 | 81 | 27 | 27 |
| 9 | 81 | 81 | 71 | 18 | 18.9 |
If we observe Sno.5 and Sno.9 closely, the earlier method would have given 18 marks to both the candidates but the new method would give the student with 79% 7.2 marks more than the person with 71% which sounds more logical because in the earlier method, a person with 71% and 79% would be treated the same but a person with 79% and 80% would not be treated so. To show this difference clearly, Sno.1 and Sno.8 would be more helpful. With the earlier method, the difference between the scores of two persons would be 18 whereas with the new method, it would be more balanced as the difference would be just about 2.7 marks. After all, the difference between Sno.1’s marks and Sno.2’s marks is 2% in each of the three biggest exams of his/her life.
Also, normalization of scores across boards and Universities would make it uniform for all the students.
Another change that could take place is some IIMs like IIM-K taking CAT scores as a parameter instead of CAT percentile because a person with 100%ile and 98%ile would have a huge difference(about 80marks) in CAT scores where as the difference is just 2 if percentiles are taken into account.
Incorporating this change would not be an issue for the IIMs as they have the latest technology holders and one of the best test organisers - Prometric conducting CAT.
2. Include new sections in CAT:
Diversity in any institute is very much required as it helps in different ideas and different mind sets interacting. But that need not be done by awarding marks straight away to one division of people and awarding 0 to the other division. If we observe why Male Engineers occupy more than 80% of the class in most of the b-schools, it could be because of the exam pattern which would be favouring them. This is only an argument and it could be wrong too. But if this is the reason that is stopping non-engineers from entering IIMs, a more balanced paper would help their cause. Adding sections like General Awareness, Decision making and Critical reasoning would make the contest more balanced. These are the sections that no specific category would have any initial edge on. More importantly, skills in the above mentioned sections are very crucial for any manager to possess. So, a CAT paper with about 5-6 sections having a decent cutoff in each section would ensure that a person with good grip in all these 5-6 sections would make it to India’s most preferred b-schools.
3. Adopt a mixed selection method:
Again, if IIMs are looking for people from diverse back grounds, then there is a better method than awarding marks to the categories who have less representation in the class. They could adopt the method that one of the premier non-IIMs adopts. They could go with a mix of profile based and score based shortlisting method. This ensures that while people with diverse profiles are given an opportunity, people who perform excellently well in CAT are not left out. They could have a minimum cut off for profile based shortlists and an exclusively CAT score based shortlist for the second type. A good mix of such kind of shortlisting would improve the diversity of the class and also ensure that CAT gets the weightage it deserves.
4. Increasing the number of calls :
Currently, IIMA, B & C adopt a 1:3 ratio in shortlisting. This could be increased to 1:4 or 1:5 to ensure that they get to select in a greater pool of students. This would require extra interview days but then, they would get to choose from a more diverse set of students.
5. Make CAT a paper-pencil exam :
There have been many cases where the student pool is unhappy with the normalization method that is adopted by the IIMs. More importantly, as the students are not given an opportunity to know the algorithm used for converting the raw CAT scores into scaled and adjusted CAT scores, there is a danger in students losing trust in the exam. This could be answered by reverting it back to the traditional paper-pencil format. This way, technical glitches which were reported in the earlier online CAT exams could be avoided. Infrastructure cannot be cited as the reason because an exam called AIEEE which was taken by 1 million students could be organized successfully in the same country with the same paper-pencil format. Also, earlier when CAT was paper-pencil based, they could conduct it successfully for 2.85 lakh aspirants and now the number is just about 2 lakhs.
6. Conducting CAT throughout the year :
If standards is what stops the IIMs from making CAT a paper based exam, they could continue with the online method but with a change. They could conduct the exam throughout the year or atleast for 6-months in a year so that more students get to take it. Also, they could allow students to take it twice or thrice in that year so that fortunes of a candidate are not decided in 2 hours. Some of the other world’s popular exams like GRE and GMAT have this system and these exams have been receiving very good response in India. Again, conducting the exam should not be a problem Prometric is the organization that takes care of GRE.
On a closing note, I wish that more than anything, if the IIMs are able to select the right set of diverse profiles ensuring that the test takers are not unhappy with the selection criteria and exam pattern, it would benefit not only the IIMs and the students but also the country on the whole.
Thank you.
- Sravanth Vangara
(The writer is a resident of Hyderabad and an electrical engineering graduate from National Institute of Technology, Warangal – Class of 2012. Currently works as a Business Technology Analyst with Deloitte Consulting. Singing and blogging are his hobbies. He subscribes to a religion called Cricket and worship Sachin Tendulkar and Rahul Dravid as Gods. Apart from this, he does cricket analytics and is also a Mathematics enthusiast. He blogs here)
You may be interested in :
India Business School Helpline - Run by IIM A C L I K
If you’re expecting calls, you must Allow us to help you! Ensure you convert that coveted call.
If you need help with your Essays, we can help you too! Prepare for GDs here
Chat
Comments
M L daga
Excellent presentation . hope, the IIM's will change their pattern in the criteria for CAT weightage & admission procedure .
21 Jan 2013, 01.51 PM
+Read Replies (1)
TheMindBlogger
Thank you :) Even I wish that the IIMs review their selection procedure and ensure that more diversity is maintained giving fair chance to every candidate :)
21 Jan 2013, 03.33 PM |
aktg47
A good report. But i have PERSONAL concerns against many of the points raised here. 1. A linear scale is not correct according to me. A simple reason is the effort required is not proportional. An increment of 2% from 63 to 65 is not equivalent to effort required to get 95% from 93%. That's why most exam results have a pyramid structure/ bell curve the way you see it. Normalizing by boards and assigning marks to Percentile scores rather than percentage scores (in any form i.e. linear/otherwise) is what is correct according to me. 2. I do not think that will make much difference. The GEMs will continue dominating high scores irrespective of what sections you include. As I have already said this in several other forums, this is a structural problem. Highest (organized) growth sectors in our Country's GDP have been IT and Manufacturing. And hence these are the sectors that have generated maximum employment after liberalization. It is natural that the brightest students in the country target these sectors and obtain the appropriate education i.e. an Engg. degree. Comparatively, our finance, education, legal, accounting or other services sectors are no where near. There has been a gradual shift in this regard in recent times with re-emergence of student interests in general universities like DU and the best students in top Engg. colleges opting for jobs in finance/ consulting . But we still have a long way to go and till then whatever exam you set, Engineers are going to ace it. But I completely agree that awarding additional marks upfront is not the solution. It's absolute disregard for meritocracy. Forced diversity according to me is going to create bigger problems in IIMs in terms of merit level of candidates. I do not have much to say about gender diversity because I do not have any analytical thoughts regarding that just opinions which I do not want to state. I agree with other points except the Paper pencil thing. The primary reason for making CAT online was to make it scalable and replicable. I think that year long CATs will soon be the order of the day. And having paper pencil format does not serve that purpose.
21 Jan 2013, 05.23 PM
+Read Replies (1)
TheMindBlogger
First of all, I appreciate the thought process behind your response. I made a point in normalization across boards. May be I must have paraphrased it better. Even my idea was to allow normalization across boards and then adopt a linear scale. Also, even if the score is not normalized, the immediate step must be to ensure that 79.9 and 80.1 do not decide selection or rejection. Also, as it an accepted fact that GEMs would be gems, we also need to try out methods which might test them. If they cross the hurdles successfully, then the whole world would have to appreciate GEMs. In case the suggested policy results in a better diversity, then the whole world would appreciate the IIMs. P.S : I'm a GEM. And I suggested paper-pencil method because it was successful for a long time and other exams like XAT, which had about 90k people attempting still follow the conventional method. Technical glitches could be avoided was what I meant. Of course, there would be other kind of problems but I believe that there were too many glitches in CAT 12 too...!!! And as you rightly suggested, year long CATs similar to GRE would make the exam attract more aspirants.
21 Jan 2013, 05.38 PM |
RTY
Ok. Good article. Kind of summarizes what most people (rather CAT givers) think. First, let us just get into the shoes of the admission committee (which includes esteemed professors who themselves were IIM Alums) and think of the situation. What IIMs want is quality of input (on which they rate the highest in the world and that too with an unbeatable record). In statistical terms what the admission committee wants to ensure is that they do not make Type II errors (and they do not care if Type I errors are made). In layman terms, the admission committee is OK to reject a candidate of good quality but they are NOT at all OK to accept a candidate of lower quality (forgive me for segregating students on "quality" basis). Thus, all the IIMs have come to be re-knowned for their stringent rules for elimination of candidates. My 2 cents worth for a healthy discussion: Point 1: One of the comments above rightly pointed that scores are not linearly proportional to efforts. So how can scoring be linear? I absolutely agree with the point you have put forward. But there is no methodology to assign weights to come up with a linear scoring mechanism. Till then, IIMs or any other institute takes in the slab process of elimination (again, strictly ensuring no Type II errors). If you agree that linear scoring is not the way (as put up by the previous comment), then all the calculations in your argument will be subject to changes. Point 2: You and I, both agree that diversity is important. We also agree that CAT is biased towards engineers. But the sample set of applicants that CAT is looking at itself has got a bias with more engineers! Where will you get the diversity from? IIMs have accepted the hard fact and going ahead with their own processes and systems to boost up chances of non GEM candidates (which, I feel is a disastrous step as Type II errors creep in). The balance has to be made such that boosting up chances of non GEM candidates do not give room for lower quality of candidates to barge in. Moreover, including more sections is the not the solutions. I'll tell you why. For one, the sample pool of applicants itself has bias, which is a natural bias. For two, testing of VA, LR and DI are hygiene factors for admission into Management Schools. General Awareness, for example, is something that is tested during the interview process. The other two sections are again covered in LR and DI part of CAT. Point 3: I do not agree with this. I feel that CAT (or any other test) must be used as an eliminating mechanism. Other factors like "profile" can be used to gauge the applicant during later processes of interviews. Point 4: Not so feasible. All the interviewers are faculty members. Imagine a 20 of them interviewing >1000 applicants. At some point, the negatives it brings in (efficiency of faculty decreases while interviewing large pool) negates the good it does. Point 5: Ah. I would love to answer this!. I have two straight objections: a) Normalisation is not a secret formula. It is a well known process that is adopted by statisticians. Everything seems transparent to me. Just that the candidates do not know how normalisation works (which he/she gets to know after they step into B-School and have a session on basic statistics). b) Years before GRE/GMAT or any other exam would have been paper based test. It is the transition which is a bumpy ride. Moreover a CBT is more efficient than a human-invigilated paper based test. Look at the decrease in cost and the increase in efficiency. CBT is believed by all IIMs that it is a powerful platform of automated system of testing that offloads a hell lot of work with human interference! c) Technical glitches: Won't you call it a problem when somewhere a person in a remote town gets an extra 5 minutes to finish off his CAT? Won't you call it a problem if a group of 3 students laid there hands on the CAT question paper the night before the exam? I believe that CBT was a strong and unanimous decision after the 2003 paper leakage fiasco. Point 6: GRE and GMAT are world wide exams. Season too change across continents - hence the year long pattern. CAT is currently a national level exam. No point in conducting it all year when the applicant pool belongs to the same geography. Multiple attempts to CAT: I again object. GRE charges you some bucks for multiple attempts, which Prometric should charge too in case of CAT being multiple-times-attempted test. This will bring a strict NO from the IIMs as there is an inherent bias towards the "affordable" in taking the test. Moreover, look at the operational issue it brings up if each of the 2 lakh applicants write the test 2 times. May be CAT may allow multiple attempts as the transition from paper based to CBT is complete, reliable and comprehensive. Last thing, it is to be understood that the faculty in the IIMs too want to teach the best students of the country. Why would they stand and witness loss of talent (if that is what has been happening, as claimed by some of the test takers). Correction will happen to make sure that the best of the talent gets in. -Personal opinions-
21 Jan 2013, 07.48 PM
+Read Replies (1)
TheMindBlogger
Well, a good response made. But, I have a few points to make. Why can't all the IIMs normalize scores of boards when some of them are able to? How would adding sections like DM and GA bring down the standards of CAT? Why are profile based shortlists not better than "free" marks schemes? Why can't ABC do it when institutes like IIM-I are able to interview 4500 people? Why have the technical glitches been continuing for the past four years where as there was a leak only once in the entire history of CAT? And why is that when an exam like AIEEE which is four times as big as CAT can do it, why can't CAT do it? And when IIMs are meant to teach Management education, why can't they do it by themselves, operate the exams for 6months or 12months and show the world that they are not just preachers but they practise what they preach?
21 Jan 2013, 11.31 PM |
ske29
Good one. . .It would be great if these changes were incorporated. 1.Conducting CAT through out the year would definitely boost up confidence, and would relieve some stress. 2. Paper and Pencil CAT will help to answer with ease.
21 Jan 2013, 07.52 PM
+Read Replies (1)
TheMindBlogger
Exactly my point!!
22 Jan 2013, 12.14 AM |
AAKRITI
a great article indeed !! and i strongly agree that iims needs to come up with some kind of changes with their stringent selection procedure. One thing in specific i would like to point is that for a commerce background it is very difficult to get a percentage of 90 or above in graduation which is one the main criterion for getting a call from iims,talking particularly about D.U ,where getting anything above 65-70 percentage is considered something really great and also first division(in annual exams not semester ) whereas compared to any engineering college getting a 9 point cgpa is something that is very common among all . Speaking about diversity how can graduate from D.U graduate if any can even be considered close to getting a call .Its Not that person from commerce background is bad at his or her subject but that's how the trend has followed in D.U that generally students are not awarded marks beyond a certain point not and perhaps that reduces his /her chances. And yes an addition to more sections like decision making,G.K and CR ,there shall be more diversification in the paper and break the typical engg. based CAT, leading to people from diverse backgrounds getting an opportunity in some ways .
21 Jan 2013, 11.55 PM
+Read Replies (1)
TheMindBlogger
Cheers to the value addition that you made Aakriti :) End of the day, every one has to get a fair chance.
22 Jan 2013, 12.12 AM |
Shrinivas
Profile based shortlisting in the first stage is not possible, given that the IIMs are answerable to RTI unlike the premier non-IIM Bschool you have pointed out. Everything will have to be made objective, which will again, end up in the point getting lost.
22 Jan 2013, 10.12 AM
+Read Replies (1)
TheMindBlogger
They could always have a quantitative weightage to the shortlist. That is precisely what they do in the interview rounds. Interview is also subjective right..!!
22 Jan 2013, 12.52 PM |
TheMindBlogger
Good point made. But, my point is that they could have a minimum cut off and then go through those profiles. This is what another premier non-IIM does.
22 Jan 2013, 03.02 PM
Anon
When I shell out 1600 bucks for CAT, I am actually a consumer. As a consumer I must know what I am buying when I pay the money. I feel conned that many of the IIMs changed there calling criteria after I filled the form. If they were to change the criteria, it should have been done before the form-filling commences. I should be entitled to the refund of my CAT form money as the criteria were changed and the changes severely hurt my chances. If I would have known I wouldnt get a call from any IIMs even after 99.4 and 1 year workex, I wouldnt have spent money on the form. Its as if the LIC has decreased the rate of Interest after you have bought a policy from them. Its unethical and illegal. And right now I am not even talking about the money that I invested in my coaching etc.
23 Jan 2013, 08.54 PM
+Read Replies (1)
TheMindBlogger
Good point made. :)
24 Jan 2013, 10.34 AM |
Chandra
At the moment CAT exam is a big farce. Person writing the exam will not know at what percentile or percentage of marks he will get call from IIMs. I doubt, if any body can derive whether he will get a call from IIMs, provided he knows his average mock test percentile (assuming his performance to be a the same level in actual exam) and his academic performance. Whole idea of normalization is statistically proven but when exam like CAT is rat race in an overly populated country like India, exam results should be worth of objective evaluation. I doubt CAT exam results and IIM calls can be put an acid test. When IIMs want profile based intake of candidates, then they should rely on proven exams like GRE( administered by Prometric) or GMAT for candidate short listing. If IIMs do not want to rely on GRE or GMAT, then they are reinventing the whole wheel causing lot of stress to lakhs of candidates every year. Its high-time for an online petition to change the CAT exam pattern. Recently XAT has to re-evaluate the XAT2013 results, when many candidates put an online petition, when they found discrepancies in XAT exam.
26 Jan 2013, 12.16 AM